CL两个塔是同一个item做两次数据增强，论文里数据增强是同时做了mask与dropout，也可以选择其它增强方式。
SSL, 20, Google自监督学习推荐算法Self-supervised Learning for Large-scale Item Recommendations 1. Motivation 在一些item特别多的...
IP属地：北京
-
-
RoBERTa, 基于BERT的优化版预训练模型
Roberta: A robustly optimized bert pretraining approachCitation: 1669 (2021-09-09) 1. M...
-
Transformer-XL，处理超级长序列的TransformerTransformer-xl: Attentive language models beyond a fixed-length contextCitation: 1326 (...
-
BERT, 18, 预训练模型Bert: Pre-training of deep bidirectional transformers for language understanding 1. Mot...
-
Transformer, NeurIPS17, Attention is all you needAttention Is All You NeedCitation: 26532 (2021-09-04) 1. Motivation 重读经典，一个重要的起点。 在作者写作...
-
MIND, CIKM19, 天猫用户多兴趣表示的召回模型Multi-interest network with dynamic routing for recommendation at TmallCitation: 52 (20...
-
ANIL, ICLR20, 重新审视MAML中的重要部分Rapid learning or feature reuse? towards understanding the effectiveness of maml. Citat...
-
-
元学习综述-introduction
Meta-Learning in Neural Networks: A SurveyCitation: 236 (2021-08-29) 1. Motivation 一个典型...
-
DRN, WWW18，深度强化学习推荐系统DRN: A Deep Reinforcement Learning Framework for News Recommendation Citation: 232 (202...
-
RMLDP, 阿里巴巴元学习上线模型Relation-aware Meta-learning for Market Segment Demand Prediction with Limited Records ...
-
SimSiam, CVPR21, KaimingHe, 自监督学习Exploring Simple Siamese Representation Learning 1. Motivation Kaiming He[1]大神的又一力作，证明在...
-
SERank，知乎基于SENet的learning to rank模型SERank: Optimize Sequencewise Learning to Rank Using Squeeze-and-Excitation Network 1. ...
-
DIN, KDD18, 阿里妈妈基于attention建模用户行为序列Deep Interest Network for Click-Through Rate Prediction 1. Motivation 本文是阿里妈妈发表在KDD18上的...
-
STAR, CIKM21, 阿里妈妈“一个模型供全部”One Model to Serve All: Star Topology Adaptive Recommender for Multi-Domain CTR Predict...
-
Meta-SSIN, SIGIR21, 元学习捕捉用户特殊兴趣Meta-Learned Specific Scenario Interest Network for User Preference Prediction 1. Motiv...
-
R-Drop, 微软两次Dropout的自监督学习R-Drop: Regularized Dropout for Neural Networks 1. Motivation 想法很直接，对模型做两次Dropout得到两个不同...